专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

1. 发明授权

EP1508893B1 Method of noise reduction using instantaneous signal-to-noise ratio as the Principal quantity for optimal estimation 有权
标题翻译：噪声降低处理使用一个信噪比主要尺寸被估计为最佳
公开(公告)号：EP1508893B1
公开(公告)日：2013-05-22
申请号：EP04103502.3
申请日：2004-07-22
申请人： MICROSOFT CORPORATION
发明人： Droppo, James G. , Deng, Li , Acero, Alejandro
IPC分类号： G10L21/02
CPC分类号： G10L21/0208
摘要： A system and method are provided that accurately estimate noise and that reduce noise in pattern recognition signals. The method and system define a mapping random variable as a function of at least a clean signal random variable and a noise random variable. A model parameter that describes at least one aspect of a distribution of values for the mapping random variable is then determined. Based on the model parameter, an estimate for the clean signal random variable is determined. Under many aspects of the present invention, the mapping random variable is a signal-to-noise ratio variable and the method and system estimate a value for the signal-to-noise ratio variable from the model parameter.

2. 发明授权

EP1693826B1 Vocal tract resonance tracking using a nonlinear predictor 有权
标题翻译：使用非线性预测跟踪声道共振
公开(公告)号：EP1693826B1
公开(公告)日：2007-08-29
申请号：EP06008561.0
申请日：2004-07-23
申请人： Microsoft Corporation
发明人： Deng, Li , Bazzi, Issam , Acero, Alejandro
IPC分类号： G10L11/00
CPC分类号： G10L25/48 , G10L25/15

3. 发明授权

EP1396845B1 Method of iterative noise estimation in a recursive framework 有权
标题翻译：一种用于在递归关系迭代噪声估计过程
公开(公告)号：EP1396845B1
公开(公告)日：2007-01-31
申请号：EP03020196.6
申请日：2003-09-05
申请人： MICROSOFT CORPORATION
发明人： Acero, Alejandro , Deng, Li , Droppo, James, G.
IPC分类号： G10L21/02
CPC分类号： G10L21/02 , G10L21/0208 , G10L21/0216

4. 发明公开

EP1465154A2 Method of speech recognition using variational inference with switching state space models 有权
标题翻译： t t t t t t t t t t t t t t t t t t t t t t t
公开(公告)号：EP1465154A2
公开(公告)日：2004-10-06
申请号：EP04007985.7
申请日：2004-04-01
申请人： MICROSOFT CORPORATION
发明人： Attias, Hagai , Lee, Leo Jingyu , Deng, Li
IPC分类号： G10L15/14
CPC分类号： G10L15/14
摘要： A method is developed which includes 1) defining a switching state space model for a continuous valued hidden production-related parameter and the observed speech acoustics, and 2) approximating a posterior probability that provides the likelihood of a sequence of the hidden production-related parameters and a sequence of speech units based on a sequence of observed input values. In approximating the posterior probability, the boundaries of the speech units are not fixed but are optimally determined. Under one embodiment, a mixture of Gaussian approximation is used. In another embodiment, an HMM posterior approximation is used.
摘要翻译：开发了一种方法，其包括：1）定义用于连续值隐藏生产相关参数和观察到的语音声学的切换状态空间模型，以及2）近似提供隐藏生产相关参数序列的可能性的后验概率以及基于观察到的输入值的序列的语音单元序列。在逼近后验概率中，语音单元的边界不是固定的，而是被最佳确定。在一个实施例中，使用高斯近似的混合。在另一个实施例中，使用HMM后验近似。

5. 发明授权

EP1199712B1 Noise reduction method 有权
标题翻译：一种用于噪声抑制方法
公开(公告)号：EP1199712B1
公开(公告)日：2009-11-25
申请号：EP01124142.9
申请日：2001-10-10
申请人： MICROSOFT CORPORATION
发明人： Deng, Li , Huang, Xuedong , Acero, Alejandro
IPC分类号： G10L21/02 , G10L15/20
CPC分类号： G10L21/0208

6. 发明公开

EP1701337A3 Method of setting posterior probability parameters for a switching state space model and method of speech recognition 有权
标题翻译：使用多Variationsinferenz与可变的状态空间模型的语音识别方法
公开(公告)号：EP1701337A3
公开(公告)日：2007-09-05
申请号：EP06013383.2
申请日：2005-01-13
申请人： Microsoft Corporation
发明人： Attias, Hagai , Lee, Leo , Deng, Li
IPC分类号： G10L15/14
CPC分类号： G10L15/14 , G10L2015/0638
摘要： A method of efficiently setting posterior probability parameters for a switching state space model begins by defining a window containing at least two but fewer than all of the frames. A separate posterior probability parameter is determined for each frame in the window. The window is then shifted sequentially from left to right in time so that it includes one or more subsequent frames in the sequence of frames. A separate posterior probability parameter is then determined for each frame in the shifted window. This method closely approximates a more rigorous solution but saves computational cost by two to three orders of magnitude. Further, a method of determining the optimal discrete state sequence in the switching state space model is invented that directly exploits the observation vector on a frame-by-frame basis and operates from left to right in time.

7. 发明授权

EP1199708B1 Noise robust pattern recognition 有权
标题翻译：噪声鲁棒模式识别
公开(公告)号：EP1199708B1
公开(公告)日：2006-11-29
申请号：EP01124141.1
申请日：2001-10-10
申请人： MICROSOFT CORPORATION
发明人： Deng, Li , Huang, Xuedong , Plumpe, Michael D.
IPC分类号： G10L15/20 , G10L15/06 , G10L21/02 , G06K9/66
CPC分类号： G10L21/0208 , G06K9/6256 , G10L15/06 , G10L15/20 , G10L21/0216

8. 发明公开

EP1557823A3 Method of setting posterior probability parameters for a switching state space model and method of speech recognition 有权
标题翻译：用于可变状态空间模型和语音识别方法的概率参数的确定方法
公开(公告)号：EP1557823A3
公开(公告)日：2005-08-24
申请号：EP05000586.7
申请日：2005-01-13
申请人： MICROSOFT CORPORATION
发明人： Attias, Hagai , Lee, Leo , Deng, Li
IPC分类号： G10L15/14
CPC分类号： G10L15/14 , G10L2015/0638
摘要： A method of efficiently setting posterior probability parameters for a switching state space model begins by defining a window containing at least two but fewer than all of the frames. A separate posterior probability parameter is determined for each frame in the window. The window is then shifted sequentially from left to right in time so that it includes one or more subsequent frames in the sequence of frames. A separate posterior probability parameter is then determined for each frame in the shifted window. This method closely approximates a more rigorous solution but saves computational cost by two to three orders of magnitude. Further, a method of determining the optimal discrete state sequence in the switching state space model is invented that directly exploits the observation vector on a frame-by-frame basis and operates from left to right in time.

9. 发明公开

EP1511007A3 Vocal tract resonance tracking using a nonlinear predictor and a target-guided temporal constraint 有权
标题翻译：跟踪使用非线性预测和目标引导的时间约束声道共振
公开(公告)号：EP1511007A3
公开(公告)日：2005-04-27
申请号：EP04103539.5
申请日：2004-07-23
申请人： MICROSOFT CORPORATION
发明人： Deng, Li , Bazzi, Issam , Acero, Alejandro
IPC分类号： G10L11/00
CPC分类号： G10L25/48 , G10L25/15
摘要： A method and apparatus map a set of vocal tract resonant frequencies, together with their corresponding bandwidths, into a simulated acoustic feature vector in the form of LPC cepstrum by calculating a separate function for each individual vocal tract resonant frequency/bandwidth and summing the result to form an element of the simulated feature vector. The simulated feature vector is applied to a model along with an input feature vector to determine a probability that the set of vocal tract resonant frequencies is present in a speech signal. Under one embodiment, the model includes a target-guided transition model that provides a probability of a vocal tract resonant frequency based on a past vocal tract resonant frequency and a target for the vocal tract resonant frequency. Under another embodiment, the phone segmentation is provided by an HMM system and is used to precisely determine which target value to use at each frame.

10. 发明公开

EP1511007A2 Vocal tract resonance tracking using a nonlinear predictor and a target-guided temporal constraint 有权
标题翻译： Verfolgen von Vokaltraktresonanzen unter Verwendung einerzielgeführtenEinschränkung
公开(公告)号：EP1511007A2
公开(公告)日：2005-03-02
申请号：EP04103539.5
申请日：2004-07-23
申请人： MICROSOFT CORPORATION
发明人： Deng, Li , Bazzi, Issam , Acero, Alejandro
IPC分类号： G10L11/00
CPC分类号： G10L25/48 , G10L25/15
摘要： A method and apparatus map a set of vocal tract resonant frequencies, together with their corresponding bandwidths, into a simulated acoustic feature vector in the form of LPC cepstrum by calculating a separate function for each individual vocal tract resonant frequency/bandwidth and summing the result to form an element of the simulated feature vector. The simulated feature vector is applied to a model along with an input feature vector to determine a probability that the set of vocal tract resonant frequencies is present in a speech signal. Under one embodiment, the model includes a target-guided transition model that provides a probability of a vocal tract resonant frequency based on a past vocal tract resonant frequency and a target for the vocal tract resonant frequency. Under another embodiment, the phone segmentation is provided by an HMM system and is used to precisely determine which target value to use at each frame.
摘要翻译：一种方法和装置将一组声道共振频率及其相应的带宽映射成LPC倒谱谱形式的模拟声学特征向量，通过为每个单独的声道谐振频率/带宽计算单独的函数并将结果相加到形成模拟特征向量的元素。将模拟特征向量与输入特征向量一起应用于模型，以确定声道谐振频率的集合存在于语音信号中的概率。在一个实施例中，该模型包括目标引导的转换模型，其基于过去的声道共振频率和声道共振频率的目标提供声道共振频率的概率。在另一个实施例中，电话分段由HMM系统提供，并且用于精确地确定在每个帧处使用哪个目标值。

你已经成功收藏专利！

检索式保存成功!

IPRDB

热门服务

关于我们

友情链接

联系方式